no place for
decontextualized bullshit testing of discrete language points
assessment used today
determine extent to which students have achieved desired results and find connection between instruction and assessment
4 basic principles that guide test creation
- test what was taught
- test in manner that reflects how it was taught
- focus test on what students can do rather than cannot do
- capture creative use of language by learners
All tests should have these 7:
- contextualized, interesting, and meaningful
- engage students in meaning-making
- elicit type of performance
- encourage divergent responses
- can be adapted to serve either as formative or summative
- address at least 1 mode of comm
- can be used or adapted to address goal areas or stds
summative
occurs at end of course, cumulative
formative
designed to help form and shape learners with opps to repair and interact
standardized
PRAXIS
proficiency
measure learner performance against criterion, ACTFL OPI
norm-referenced
standardized and measured against pop, SAT
criterion-reference
classify indiv according to whether they can perform a certain task
validity
learners can expect that test will measure what was taught
reliability
test will be graded objectively as if it didn’t matter who graded it, they will still get about the same grade
PBA
performance based assessment, create a product or response based on knowledge or skills
IPA
integrated performance assessment measure student progress in attaining competencies described in natl stds, provide ample practice time, activities must be linked, cannot do 4 w/o doing 1 2 and 3
